PromptPrivacy – AI Privacy & PII Protection
An AI privacy solution that detects and removes personal or sensitive information from prompts before they are sent to large language models.
Project Overview
PromptPrivacy is an AI privacy solution designed to protect sensitive information before it reaches large language models. We worked on a system that analyzes user prompts, identifies personal or sensitive data, and applies masking or redaction before forwarding the request to the AI provider. The platform combines rule-based detection with named entity recognition to provide an additional privacy layer for organizations adopting generative AI.
Key Features
Technical Challenges
Detecting sensitive information without removing important prompt context
Balancing privacy protection with the quality of AI responses
Supporting different types of personal and organization-specific data
Designing configurable filtering rules for different use cases
Results & Impact
Project Info
PromptPrivacy
1–2 months
—
2023
Technologies Used
Project Links
Project Gallery
